Book Review
Syst. Biol. 53(5):835–838, 2004
Copyright c© Society of Systematic Biologists
ISSN: 1063-5157 print / 1076-836X online
DOI: 10.1080/10635150490522322
Gentianaceae: Systematics and Natural History.—Lena Struwe and
Victor A. Albert (eds.). 2002. Cambridge University Press, Cambridge,
CB2 2RU, UK. x + 652 pp. ISBN 0-521-80999-1. Hardback. £95 ($130).
We are witnessing a revolution in plant systematics
brought on by the flourishing of molecular phylogenet-
ics principally based on DNA sequence data. Over the
last decade, thousands of molecular phylogenies have
been published for all types of organisms including al-
most all major groups of plants, addressing overall rela-
tionships among major groups or detailed relationships
in specific lineages (Pagel, 1999). In addition to studies
on the evolution of genes, genomes, and the origin of
biodiversity, recent advances have led to the assembly
of the general ”tree of life” and to the compilation of
phylogenetic trees into hierarchical systems, i.e., clas-
sifications. Such practice is done at or above the level
of family, e.g., the work of the Angiosperm Phylogeny
Group (APG, 2003). Gentianaceae: Systematics and Natural
History combines molecular and morphological data to
revise the outdated classification of a family established
more than a century ago. Evolutionary trends in mor-
phology, anatomy, palynology, and phytochemistry are
reconsidered in the light of molecular phylogeny.
The first chapter, coauthored by the editors, is a con-
cise overview of the Gentianaceae: its taxonomic history,
circumscription, diagnostic characters, affinities and dis-
tribution. Based on a review of phylogenetic studies at
the family level or above, family Gentianaceae of Gilg
(1895) is modified to include Anthocleista, Fagraea and
Potalia (formerly Loganiaceae), and Saccifolium (formerly
recognized as a monotypic family), but excluding sub-
family Menyanthoideae, which has been confirmed as a
family of Asterales by several recent phylogenetic stud-
ies. Gentianaceae so defined upholds an increased mor-
phological heterogeneity and, as the authors indicate,
lacks a general characterization, but its monophyly is
well supported by DNA data and potentially by some
cryptic phytochemical features. The authors attempt to
characterize the newly defined Gentianaceae, but the de-
scription of several general morphological gestalts seems
rather ineffective in this regard, although highlighting
some evolutionary problems in the family. It would have
been useful, particularly for those who do not know
much about the Gentianaceae, if the authors had pre-
sented a comparison with closely related families of the
order Gentianales. Statistical descriptions of diversity in
terms of the number of species and genera are also pre-
sented for major clades and partitions of the continents.
Although a general description of diversity and its geo-
graphic distribution is necessary, the figures can be mis-
leading, particularly concerning the number of genera
in a clade or a region, because generic delimitations of
some clades are highly controversial.
Chapter 2, coauthored by Struwe et al., is the most sig-
nificant part of the book, in terms of both length (it covers
nearly half of the volume) and scientific importance, pre-
senting a new classification integrating molecular phy-
logeny with morphology. Since Gilg (1895), the classifi-
cation of Gentianaceae has been somewhat overlooked,
and revision of the classification was “a work waiting
to be done.” Phylogenetic relationships inferred from
two plastid fragments, the trnL(UAA) intron and matK
gene, were used as the backbone of the new classifica-
tion. The molecular phylogenies were produced mainly
with previously published data with a few additions of
taxa. This chapter primarily provides detailed descrip-
tions of the tribes, subtribes, and genera, with morpho-
logical evolution and biogeography discussed in the con-
text of molecular phylogenies. Although providing an
adequate guideline for classification, discussion on phy-
logenetics per se falls short and is sketchy in the book. It
would have been better to justify the new classification
with a thorough discussion of the phylogenetics of the
family. Nevertheless, the inferred phylogenies are con-
gruent with those published earlier (Thiv et al., 1999), and
in full agreement with our recent results based on both
plastid and nuclear DNA sequences in terms of tribal and
subtribal relationships (Yuan et al., 2003). The phyloge-
netic assumption of the classification, at least at the tribe
and subtribe levels, seems solid and well supported.
The tribe Saccifolieae is newly established by Struwe
et al. in this chapter to accommodate a basal clade consist-
ing of several small genera that were previously placed
elsewhere both inside and outside Gentianaceae. The
recognition of this small tribe significantly enlarges the
morphological breadth of Gentianaceae, such as sac-
cate leaves and heterostyly, and has great importance
in understanding the origin and early differentiation of
Gentianaceae.
The tribe Exaceae, although not new, is newly shaped
in Chapter 2 to include the Madagascan endemics Tachi-
adenus and Gentianothamnus. Dr. Klackenberg, author
of this tribe, has devoted study to this tribe since his
doctoral thesis, a monograph on the genus Exacum. His
descriptions of morphological, anatomical, and kary-
ological variations, as well as phylogenetic relationships
among genera of the tribe, are by far the most detailed
835
at U
niversity of Reading on N
ovem
ber 2, 2011
http://sysbio.oxfordjournals.org/
D
ow
nloaded from

836 SYSTEMATIC BIOLOGY VOL. 53
and sophisticated of the tribes presented in the book. His
suggestions of a close affinity between the saprophytic
genus Cotylanthera and Exacum based on morphology
(anthers opening by apical pores and finely perforated
endothecial walls), and close relationship between Gen-
tianothamnus and Tachiadenus, are all confirmed by our
recent molecular studies (Yuan et al., 2003). Meanwhile,
he suggested gondwanan tectonic history as the cause
of the vicariant distribution pattern of the deep branch-
ing clades of Gentianaceae, within the tribe Exaceae, and
even the genus Exacum. He further hypothesized that the
genus Exacum may have existed over 80 million years be-
fore present (Mya). These speculations are, however, not
corroborated by molecular dating, because divergence
of the order Gentianales is estimated at not beyond the
Late Cretaceous (< 90 Mya) and divergence of the family
Gentianaceae at around 50 Mya (Wikstro¨m et al., 2001;
Yuan et al., 2003).
Following molecular phylogenetic results (Thiv et al.,
1999), the tribe Chironieae is newly defined in this chap-
ter to accommodate Gilg’s (1895) subtribes Erythraeinae
and Chironiinae. The tribe is further divided into three
subtribes that seem to show biogeographic importance.
Excepting the delimitation of some genera such as
Centaurium, the proposed circumscription of tribe and
subdivision into subtribes is supported by independent
studies (Mansion, 2001). The suggested biogeographic
histories of the tribe and the subtribes concur with
molecular dating (Wikstro¨m et al., 2001; Mansion, 2001).
Descriptions and discussion of this tribe fall short, but
subsequent publications by the authors provide a useful
addition (Thiv and Kadereit, 2002; Thiv, 2003).
The classification of the tribe Helieae presented in this
chapter, rooted in the thorough studies of the doctoral
thesis of Struwe (1999), has greatly clarified the long-
lasting taxonomic and nomenclatural confusion regard-
ing the genera of the group. The genera of this tribe
used to be placed in three different tribes by Gilg (1895)
based mainly on the superficial similarity of pollen mor-
phology. Although the importance of pollen characters
to infratribal classification of Helieae was confirmed,
molecular phylogenetic results have shown that pollen
types, viz. monads, tetrads, and polyads, are superfi-
cially homoplasious, and involve substantial difference.
For example, tetrad and polyad pollen were the diagnos-
tic character of Gilg’s (1895) Helieae, thus genera such
as Coutoubea, Deianira, Schultesia, Symphyllophyton, and
Xestaea, all producing tetrads, were included in Helieae
by Gilg, whereas molecular results revealed them to
be members of Chironieae. Enlightened by molecular
studies, further palynological observations on these gen-
era confirmed that the tetrads of these genera were in
fact significantly different from those of typical Helieae,
such as Adenolisianthus, Helia, Tetrapollinia. The circum-
scription of the tribe Helieae and the genera therein are
better balanced between palynological data and other
morphological data. The genera in the Helieae of the new
classification are small and much divided except genus
Macrocarpaea (Grant, 2003), which is not relevant to some
other tribes, such as Gentianeae.
The tribe Potalieae, although the name per se and its
Gentianaceae attribution is not new, is newly shaped
in this chapter to comprise three subtribes. Molecular
data, strongly supporting the Gentianaceae affiliation of
the subtribes, conflict with morphological data includ-
ing pollen and seed micromorphology revealed in this
book, which deserve further studies. The morphologi-
cal descriptions and evolutionary considerations seem
relevant and detailed. Nevertheless, the biogeographic
suggestions are not convincing. For example, it is hard
to believe that the current distribution of Potalia elegans
in the white-sand of the Guayana Shield is relictual dat-
ing from gondwanan continuity between South America,
Africa, and Madagascar as the authors suggest (p. 206).
It is equally hard to believe that the repeated disjunct
pantropical patterns shown in the subtribe Potaliinae
and the species and subspecies of the genus Enicostema
were due to ’a common vicariance event’ (pp. 219–220).
A common vicariant origin of the patterns would suggest
unrealistically old ages for the subtribe and the genus (or
even the subspecies of Enicostema axillare), and also a dra-
matic slow down of evolution in the annual or perennial
herb genus Enicostema, or dramatic speed up of evolution
in the woody mostly tree subtribe Potaliinae consisting
of Anthocleista, Fagraea, and Potalia. The repeated disjunc-
tive patterns shown in the subtribe are not likely to have
a common cause.
The tribe Gentianeae consists of mostly temperate gen-
tians that are better known and studied. The tribe Gen-
tianeae of the new system presented in this chapter
is principally congruent with subtribe Gentianinae of
Gilg (1895) with the inclusion of Bartonia and Obolaria
and exclusion of Ixanthus. Two evolutionary lineages,
readily revealed in the tribe Gentianeae by both morpho-
logical and molecular studies, are properly accommo-
dated as two subtribes in the new system, Gentianinae
and Swertiinae. Except for the extremely heterogeneous
genus Swertia, and to a lesser extent Gentianella and Lo-
matogonium, most recent authors of Gentianaceae tend to
agree on the circumscriptions and relationships among
the most genera of the tribe. Swertia as currently defined
is highly polyphyletic containing at least 14 lineages with
regards to other genera of the subtribe as revealed by a
recent phylogenetic studies with extensively sampling
of species worldwide and multiple molecular markers
of both nuclear and plastid genomes (Chassot, 2003).
Further taxonomic and nomenclatural adjustments are
necessary and anticipated. Although biogeography is
still inconclusive, the diversification pattern of the tribe
across the entire northern hemisphere does offer a good
case to study the extensive floristic links between the
Eurasian and American continents.
In general, the new classification has revealed
significant taxonomic clarity, and has improved our un-
derstanding of the general relationships within the Gen-
tianaceae. It will serve as a good platform for further
studies at lower taxonomic levels, yet we still have to
decide on a good genus concept that can be applied
to all tribes of the family. Apart from the less under-
stood polyphyletic genera (e.g., Centaurium, Gentianella,
at U
niversity of Reading on N
ovem
ber 2, 2011
http://sysbio.oxfordjournals.org/
D
ow
nloaded from